The Agent Skills Directory

[PROMPT_INJECTION]: No evidence of prompt injection, jailbreak attempts, or instructions to bypass safety guardrails was found. The rules defined in the skill are operational constraints related to the synthesis workflow (e.g., citing claims, following the '3x rule').
[DATA_EXFILTRATION]: No hardcoded credentials, sensitive file path access, or network operations (such as curl or wget) were detected. The skill focuses on processing local note content within a predefined workspace.
[REMOTE_CODE_EXECUTION]: The skill does not perform any external downloads, package installations, or execution of remote scripts.
[INDIRECT_PROMPT_INJECTION]: The skill includes a data ingestion surface by processing notes from a user's vault. While there are no explicit boundary markers or sanitization steps mentioned to ignore embedded instructions in those notes, the skill does not utilize any high-risk tools (like network access or arbitrary command execution) that could be abused via malicious content in the notes.
[OBFUSCATION]: No Base64, zero-width characters, homoglyphs, or other obfuscation techniques designed to hide malicious intent were identified in the instructions or metadata.

reflect